JP2018533999A - Ventilator with error detector for flow sensor - Google Patents

Abstract

  The ventilator (10) for performing artificial respiration according to the present invention has in particular a flow sensor assembly (44, 48) for quantitatively detecting the gas flow in the artificial respiration conduit assembly (30). The flow sensor assembly includes a distal flow sensor (48) disposed remotely from a patient end of the ventilator conduit assembly (30) and a patient end of the ventilator conduit assembly (30). The ventilator (10) has at least a controller (18) for processing the measurement signals of the flow sensor assembly (44, 48). The controller (18) is configured to infer errors based on measurement signals of the distal flow sensor (48) and / or the proximal flow sensor (44). According to the present invention, the control device (18) allows the change values (62, 76) of the measurement signals (54, 58, 68, 72) of one flow sensor (44) to be respectively detected by the other flow sensor (48). For comparison with the change value (60, 74) of the measurement signal (52, 56, 66, 70) and / or comparison with the measurement signal (54, 58, 68, 72) of one flow sensor (44) Based on this, the flow sensor assembly (44, 48) is configured to infer errors.

Description

The present invention relates to a respirator for performing artificial respiration, the respirator,
-Artificial ventilation gas source,
-A ventilator conduit assembly extending between the source of ventilator gas and the proximal end on the patient side;
-A valve assembly including an inhalation valve and an exhalation valve;
A flow sensor assembly for quantitative detection of gas flow in the ventilator conduit assembly, the distal flow sensor being located far from the patient end of the ventilator conduit assembly; and the ventilator conduit assembly A flow sensor assembly having a proximal flow sensor located near the patient end, and
A controller for processing at least the measurement signal of the flow sensor assembly, the controller configured to infer an error based on the measurement signal of the distal and / or proximal sensor;
have.

  As this type of respirator, for example, a commercially available product called “SERVO-U” manufactured by Maquet Corporation is known. This known ventilator uses an internal sensor present inside the ventilator as a distal flow sensor. Connected to the ventilator is the distal end of the ventilator conduit assembly. Furthermore, this known ventilator uses a proximal flow sensor in the form of a hot-wire anemometer in the Y-piece, within the proximal flow sensor, in the direction towards the patient, an inhalation hose and an expiration hose And integrated into a common ventilator conduit extending toward the patient. According to this known ventilator instruction manual, the output of the internal pressure sensor and flow sensor is compared with the measurement result of the proximal sensor in the Y-piece, and is important between the values used for comparison. If there is a significant difference, the proximal sensor is turned off.

  The drawbacks of known ventilators and the error detection that becomes apparent are that error detection based on the comparison of the measurements of two different sensors is only part of the error that is actually occurring in the sensor assembly. It is because it cannot be detected. In addition, an artificial respiration conduit assembly is present between the two sensors in the form of an elastic hose that is counteracted by the breathing gas that passes through the elastic hose with each breath. And stretched. Based on the elasticity of this hose, the two flow sensors placed at different positions of the breathing gas flow provide different measurements even at the same effective breathing gas flow at both detection positions. That is, the breathing gas that stretches the hose passes through the distal flow sensor but not through the proximal flow sensor, so the distal flow sensor reading is generally higher than the proximal flow sensor reading.

  A ventilator is an important device for sustaining life for patients who are unable to breathe spontaneously or who cannot fully breathe spontaneously. Thus, its exact function is critical. The exact function of the ventilator again depends on the most accurate detection of the amount of breathing gas delivered to the patient.

German Patent Application No. 102010040287

  An object of the present invention is to further develop a ventilator of the type mentioned at the beginning so that an error can be recognized early and reliably when an error occurs in the function of the flow sensor assembly.

  According to the present invention, this problem is solved by an associated ventilator. In the ventilator, the control device is configured to change the measurement value of the measurement signal of one of the proximal flow sensor and the distal flow sensor, and the flow of the other of the proximal flow sensor and the distal flow sensor. An error of the flow sensor assembly is configured to be inferred based on the comparison with the change value of the sensor measurement signal. By using the change value of the measurement signal instead of the measurement signal itself, the error of the flow sensor assembly can be detected earlier and / or reliably than before. Thus, for example, if the proximal flow sensor reading and the distal flow sensor reading are changing slightly but in different directions, for example, one flow sensor indicates an increase in flow rate. If each other flow sensor indicates a decrease in flow, it is already possible to detect an error. Therefore, it is already possible to infer an error even if the measurement values of the two flow sensors are compared with each other and no significant difference is shown.

  Alternatively or additionally, the change value of the measurement signal of one of the proximal flow sensor and the distal flow sensor is taken as the measurement signal of the same flow sensor, i.e. again, the measurement of this one flow sensor. If the change value for the latest measurement signal before the change based on the prevailing operating conditions is abnormally large or smaller, for example, one and the same flow sensor, i.e. Based on only the signal of the distal flow sensor or the distal flow sensor, the error of this one flow sensor can be inferred without considering the signal of the other flow sensor.

  In this case, when comparing change values of different sensors with each other, simultaneous change values or almost simultaneous change values, that is, change values detected at the same time or in the same breath are important.

  When the change value of the measurement signal is compared with the measurement signal, the change value is compared with the measurement signal detected earlier in time, preferably with the measurement signal detected immediately before the change value in time. is important.

  The solution proposed by the present invention therefore makes it possible to detect errors on a larger scale than is known for prior art ventilators, in addition to the error of one flow sensor, It is possible to detect regardless of the signal of the other flow sensor.

  Particularly reliable error recognition is obtained when the above-described possibilities for detecting errors in the flow sensor assembly are used not only alternatively but also redundantly.

  As already suggested, fast and effective error detection means that the change value of the measurement signal of one of the distal flow sensor and the proximal flow sensor is opposite to the change value of the other flow sensor. This is possible if the controller is configured to infer an error in the flow sensor assembly when it has a different direction of change. At that time, in order to estimate the error, it is not necessary to compare the numerical values with each other, and it may be sufficient to compare the signs of the change values.

  According to previous experience, there are mainly two different types of errors in flow sensor assemblies, especially proximal flow sensors. These types of errors differ mainly throughout the period of development and action on the ventilator.

  In the following, the error denoted as “spring error” appears suddenly and causes a serious error in the measurement result within a few seconds.

  The other error, referred to below as “drift error”, appears gradually and leads to serious errors in the measurement results over a few minutes up to 30 or 45 minutes.

  Based on previous recognition, these two errors are most likely related to the condensate precipitation in the proximal flow sensor. By providing the proximal flow sensor very close to the patient, the proximal flow sensor may be more sensitive to moisture in the breathing air. While a spring error is very well understood, particularly in a flow sensor that works on the principle of pressure differential, with respect to drift error, a flow sensor that has become cloudy with condensate after a drift error has been detected and a dry flow sensor. Until now, it was only possible to confirm that the drift error disappears when replaced.

  Even though the at least one proximal flow sensor is preferably a flow sensor that works on the principle of pressure difference, as is known, for example, from US Pat. It should not be limited to a ventilator having a flow sensor assembly with at least one flow sensor. The improved ventilator proposed here has improved the accuracy of error detection in the flow sensor, and basically the flow sensor uses any physical operating principle to measure the gas flow. Any arbitrary flow sensor can be used regardless of whether it is used.

  It should be mentioned for completeness that, according to current theory, spring errors are continuously detected in the direction of flow within the sensor in the proximal flow sensor operating on the pressure differential principle described in US Pat. A movable flap that separates the two pressure measurement chambers that are placed is immersed in the condensate that has accumulated in the sensor, which, due to its movement, provides greater resistance than in a dry sensor. Occurs when it must be overcome.

Spring errors can advantageously be detected quickly and reliably if the control device is configured to infer errors in the flow sensor assembly in the following cases.
The ratio of the total change value of the measurement signal of the proximal flow sensor to the first reference value exceeds a predetermined reference threshold; and
The ratio of the total change value of the measurement signal of the distal flow sensor, which occurs before and after the second reference value, falls below a predetermined reference threshold value.

  In this case, that is, the change values occurring in succession of the measurement signals of the proximal flow sensor and the distal flow sensor are summed up, so that the change in the measurement signal of each flow sensor over the total period of time is summed. The whole is quantified. This total change in measurement signal can be checked against a reference value to assess whether the flow sensor assembly is functioning correctly. This evaluation can be performed using a predetermined reference threshold.

  Preferably, the change values of the measurement signal that occur directly in succession are summed to form a total value, thereby forming a total value over the shortest possible period, and within the flow sensor assembly It is possible to quickly detect the corresponding error. Furthermore, it can be ensured that there are no measured signal change values that are not taken into account for the evaluation of the accuracy of the operation of the flow sensor assembly.

  According to a preferred further development of the invention, the first and / or second reference value is a measurement signal value of the proximal flow sensor. Thereby, the change value of the proximal flow sensor and the change value of the distal flow sensor are respectively changed in the change of the proximal flow sensor that is an error-prone flow sensor of the proximal flow sensor and the distal flow sensor. It is possible to check against the total. This is often useful for evaluating the correct operation of the flow sensor assembly. This is because the absolute value of the change in the measurement signal can generally be effectively evaluated with respect to its validity only in the context of measurement values that are changed through the change. Correspondingly, the accuracy of error detection, in particular spring error detection, is determined by the measured signal value of the proximal flow sensor when the first and / or second reference value is at or just before the start of the sum of change values. It can be improved by being. These changes are directly related to the measurement signal that is the object of verification for the assessment of its validity or accuracy.

  However, the ventilator does not need to perform error detection continuously throughout the duration of its operation. If this need becomes apparent from ventilator operation, it is sufficient to provide the computing power of the ventilator controller for more accurate error detection.

  Thus, preferably, the ventilator controller is configured to continuously monitor the operation of the flow sensor assembly for errors according to a first, less computationally demanding test process; Further, if the first test process indicates a suspected state of the flow sensor assembly, the operation of the flow sensor assembly is subject to error according to a second, more computationally intensive test process. Configured to continuously monitor for.

  This is so that, for the preferred case of the intensive testing process described above, the controller starts summing the change values when the measured signal value and / or change value of the proximal flow sensor exceeds a predetermined signal threshold. It can mean that it is configured. The test of whether the measured signal value and / or change value of the proximal flow sensor is above a predetermined signal threshold is preferably carried out continuously and requires a less computational power as described above abstractly. 1 test process. If the change value of the proximal flow sensor and / or the measured signal value is above a predetermined signal threshold, a suspected error condition has been detected, preferably the sum of the change values described above and its reference value The comparison is started.

  In parallel, it can be considered that the first test process is continued and the second test process is terminated again, because the first test process is suspected of being in error for a predetermined number of breathing cycles. When the measured signal value and / or change value of the proximal flow sensor no longer exceeds the predetermined signal threshold over a predetermined number of continuously evaluated measurement signals, for example. .

  However, it is also conceivable to continue the second, more computationally intensive test process until the end of ventilator operation once a suspected error condition has been detected.

  In a preferred embodiment of the ventilator, the distal flow sensor is used for the most accurate control of breathing gas flow while breathing. This means that the breathing gas flow is altered based on the measurement signal supplied by the distal flow sensor while breathing.

  In contrast, proximal flow sensors can be used to control minute ventilation, as used, for example, in known ventilation modes ASV, Intelligent-ASV and APV. In so doing, the minute ventilation is preferably adjusted or controlled for each breath. That is, based on the measurement signal of the proximal flow sensor, the breathing gas flow is not changed while breathing. This has the advantage that the amount of breathing gas actually delivered to the patient can be determined more accurately than with a distal flow sensor using a proximal flow sensor located near the patient. Especially because effects that negatively affect the accuracy of the measurement signal, such as the elasticity of the hose conduit in artificial respiration hoses, do not affect the measurement signal of the proximal flow sensor or, if at all, have a very small scale Because it is only.

  Because of the arrangement of the distal flow sensor and the proximal flow sensor, the distal flow sensor preferably allows exclusively inspiratory gas flow, while the proximal flow sensor allows both inspiratory and expiratory gas flow. Again, it is clear that the proximal flow sensor is more prone to errors than the distal flow sensor.

  Even if the ventilator changes the breathing gas flow while breathing, as described above, basically, since the patient is ventilated with discontinuous breathing, preferably, the controller is It is also conceivable that the measurement signals of the proximal flow sensor and the distal flow sensor are processed for each breath. This has the advantage that most ventilation parameters that determine the operation of the ventilator exist for each breath, ie for each breath to be performed.

  In this case, according to a preferred further development for further increasing the accuracy of error detection, the first and / or second if the measured signal value of the proximal flow sensor exceeds a predetermined signal threshold value in threshold breathing. It is conceivable that the reference value is the measurement signal value of the proximal flow sensor assigned to the breath immediately before the threshold breath. This also forms a direct temporal relationship as much as possible between the measurement signal and / or change value to be used for error detection of the flow sensor assembly and the reference value against which the signal or value is to be matched. Used for.

Experiments show that the drift error is when the controller is:
The ratio of the total change of the measurement signal of the proximal flow sensor, which occurs in succession, preferably in direct succession, to a third reference value exceeds a predetermined reference threshold; and
If the ratio of the total change value of the measurement signal of the distal flow sensor, which occurs in succession, preferably in direct succession, to the fourth reference value is below a predetermined reference threshold value,
In order to guess the error of the flow sensor assembly,
Or if:
The ratio of the proximal flow sensor measurement signal, which occurs in succession, preferably in direct succession, to the fifth reference value, below a predetermined reference threshold, and
If the ratio of the sum of the change values of the measurement signal of the distal flow sensor, which occur one after the other, preferably in direct succession, to a sixth reference value exceeds a predetermined reference threshold value,
Can be detected well by being configured to infer errors in the flow sensor assembly.

  It should be clarified here that in this application the use of ordinal numbers for distinguishing the reference threshold represents the order in which the reference threshold is listed in the application document, and the minimum of the reference threshold in the ventilator It should not be understood as a numerical description.

  With regard to the advantages of using directly continuous change values, reference is made to the above description regarding the detection of spring errors, which is valid here as well. The listed reference thresholds have been calculated by experimentation, taking into account medical boundary conditions, without any significant costs.

Basically, the drift error can already be reliably detected under the above-mentioned conditions. Higher accuracy of drift error detection is obtained by the controller being configured to infer errors in the flow sensor assembly only in the following additional cases.
The sum of the change values of the measurement signal of the proximal flow sensor that occur one after the other, preferably in direct succession, exceeds a predetermined first sum threshold, and / or
The sum of the change values of the measurement signal of the distal flow sensor, which occur one after the other, preferably in direct succession, exceeds a predetermined second total threshold value.

  Again, redundant application of the above conditions increases accuracy. For both sensors, the summation is preferably performed from a common starting point to each last breath.

  Basically, any reference value that appears to be appropriate can be used to evaluate the validity or accuracy of the change value of one of the two flow sensors or the measurement signal of the two flow sensors. is there. As already indicated, a particularly meaningful reference value for evaluating the change of the measurement signal or the measurement signal itself is the measurement signal value of the proximal flow sensor. Thus, in an advantageous further development of the invention, it is defined that the third and / or fifth reference value is a measurement signal value of the proximal flow sensor. In order to create as direct a temporal relationship as possible between the reference value and the variable to be evaluated, the third and / or fifth reference value is preferably at the beginning of the sum of the change values or It is a measurement signal value of the proximal flow sensor just before the start.

  With the same motivation, in a preferred further development of the invention, the fourth and / or sixth reference value is the measured signal value of the distal flow sensor, in particular the distal flow at the beginning or just before the start of the sum of the change values. It is specified to be a measurement signal value of the sensor.

  Again, it is specified that the sum of the change values to form a sum value starts when the value of the change value of the measurement signal of the proximal and / or distal flow sensor exceeds a predetermined threshold value. obtain.

  Again, the comparison of the change value of the measurement signal of the proximal flow sensor and / or the distal flow sensor with the predetermined threshold value is less computational power in the first sense above. And starting from that test process can result in a test process that requires the second, more computational power described above. With regard to the duration of the test process that requires more computational power to detect drift errors, reference is made to the above description regarding detection of spring errors, which again relates to detection of drift errors. It is effective.

  A first test process for determining whether to initiate a second, more computationally intensive test process may alternatively or additionally involve proximal and distal flows, respectively, for the same breath It may be related to the direction of change of the sensor change value, i.e. for example the comparison of signs. If the signs are different, i.e. the measurement signals of the proximal and distal flow sensors change in different directions for the same breath, there is suspected error sufficient to start the second test process doing.

  The control device can be configured to smooth the measurement signal of the flow sensor by filtering so as not to include as much of the obstruction as possible in the consideration of the variables used for error detection. Such an inhibiting factor may be spontaneous breathing of a patient who is ventilated using, for example, a ventilator.

  A possible type of filtering is the formation of a moving average. Since the spring error occurs fast, i.e. within a few seconds, it should form a moving average for several change values and / or measured signal values, e.g. not exceeding 25, preferably not exceeding 20, e.g. exceeding 16. It is enough. Preferably, a moving weighted average may be formed, for example, to weight differently breaths taken at different times. For example, when considering a moving average, it may be meaningful not to consider more recent values for the latest measurement than values that go further in time. Similarly, the value to be filtered at the end of the time frame to be observed for filtering may be weighted less than, for example, the value located in the middle of the time frame to be observed for filtering, It can positively affect the accuracy of error detection.

  With regard to the determination of the drift error, the measured signal values of both flow sensors can each be smoothed by a moving average, in particular a weighted moving average.

  Additionally or alternatively, the measured signal value to be smoothed can also be filtered by a digital filter having an order higher than 1, and by filtering with an order higher than 2, particularly preferably a fourth order digital By using the filter, a more excellent smoothing effect can be obtained without losing accuracy in error detection. Such a digital filter is preferably used in detecting a drift error.

  Drift errors appear gradually over a relatively long period of several minutes, so smoothing of the measurement signal with a weighted moving average or digital filter as necessary is preferably smooth for detecting spring errors. A larger number of measurement signals are used than in the case of conversion. Preferably, for smoothing for drift error detection, at least the latest 50 measurements, particularly preferably at least the latest 75 measurements, most preferably at least the latest 100 measurements are used.

  The control device is preferably configured to detect the change value based on the smoothed measurement signal.

  It should be clarified here that a recursive filter rule using a filter coefficient a according to the following structure is understood as a first-order digital filter.

  Here, y (n) is a filter output related to the nth breath, y (n-1) is a filter output related to the most recent (n-1) th breath, and x (n) is n Is the latest input signal for the second breath, and a is the filter coefficient determined for the filter.

  In so doing, the m-th order filter is applied m times as much as the above filter structure with the i-th order filter output for the n th breath as the input signal of the next higher (i + 1) th order filter for the n th breath. In this case, 1 ≦ i ≦ m is effective.

  Preferably, the proximal flow sensor and the distal flow sensor function discretely in time rather than continuously, so the proximal flow sensor and the distal flow sensor are each assigned to that breath for each breath. One, preferably just one measurement value representing the gas flow is supplied.

  In order to be able to derive measures from error detection in addition to detecting an error, the control device is preferably configured to issue an alarm when an error is inferred. However, unlike the prior art, it is not necessary to turn off the sensor. Accordingly, the controller is preferably configured to continue processing the measurement signal of the proximal flow sensor, and particularly to investigate the presence of a sensor error.

  Hereinafter, the present invention will be described in detail with reference to arbitrary drawings. The following figure is shown.

It is the figure which showed embodiment which concerns on this invention of the ventilator which concerns on this invention. FIG. 5 shows raw measurement signals and smoothed measurement signals for a distal flow sensor and a proximal flow sensor. It is a figure which shows the change value of the measurement signal of a distal flow sensor and a proximal flow sensor. FIG. 5 is a further diagram illustrating raw and smoothed measurement signals of the distal flow sensor and the proximal flow sensor. FIG. 6 is a further diagram showing the change values of the measurement signals of the distal flow sensor and the proximal flow sensor. FIG. 5 is a very schematic diagram relating to error detection of a spring error. It is a very schematic diagram regarding error detection of drift errors.

  In FIG. 1, an embodiment of a ventilator according to the present invention is indicated by reference numeral 10 as a whole. In the illustrated example, the ventilator 10 is used for artificial ventilation of the patient 12.

  It should be pointed out merely for completeness that the ventilator 10 according to the present invention may be received as a movable ventilator 10 on a rotatable table 13.

  The ventilator 10 has a housing 14, and the pressure change device 16 and the control device 18 may be received in the housing 14 because the housing material is opaque and cannot be visually recognized from the outside.

  The pressure change device 16 is configured as known and may include a pump, a compressor, a blower, a pressure accumulator, a pressure reducing valve, and the like. Furthermore, the ventilator 10 has an inhalation valve 20 and an exhalation valve 22 as is known.

  The control device 18 is generally realized as a computer or a microprocessor. The control device 18 includes a storage device not shown in FIG. 1 so that data necessary for the operation of the ventilator 10 can be stored and recalled when necessary. The storage device may be located outside the housing 14 when the network is operated, and may be connected to the control device 18 by a data transmission connection unit. The data transmission connection portion may be formed by a cable path or a transmission path. However, in order to prevent a failure of the data transmission connection from affecting the operation of the ventilator 10, the storage device is preferably integrated into the control device 18, or at least in the same housing as the control device 18. 14 is received.

  In order to input data to the ventilator 10 or more precisely to the control device 18, the ventilator 10 has a data input 24, which is shown in FIG. In the example, it is represented by a keyboard. The control device 18 may obtain data through various data input units such as a network line, a transmission line, or the sensor connection unit 26 instead of or in addition to the illustrated keyboard. These various data inputs will be discussed in more detail below.

  The ventilator 10 can have an output device 28 for outputting data to the treating clinician. In the illustrated example, the output device 28 is a screen.

  For ventilation, the patient 12 is connected through the ventilator assembly 10, more precisely the pressure change device 16 in the housing 14, and the ventilator conduit assembly 30. For this purpose, the patient 12 is intubated.

  The artificial respiration conduit assembly 30 has an inhalation hose 32 through which fresh breathing gas can be directed from the pressure change device 16 to the lungs of the patient 12. The intake hose 32 can be interrupted and can have a first intake partial hose 34 and a second intake partial hose 36 between which to intentionally add moisture and If necessary, an adjustment device 38 for adjusting the temperature of fresh breathing gas supplied to the patient 12 can be provided. The adjustment device 38 can be connected to an external liquid container 40, and the liquid container 40 is used to supply water for supplying moisture, or for example, a medicine for suppressing inflammation or dilating the airway. Can be supplied. When the ventilator 10 according to the present invention is used as an anesthesia ventilator, a volatile anesthetic can be managed by the ventilator 10 and given to the patient 12. The adjustment device 38 adjusts so that fresh breathing gas is supplied to the patient 12 at a predetermined humidity at a predetermined humidity and, optionally, a pharmaceutical aerosol.

  The artificial respiration conduit assembly 30 has an exhalation valve 22 and an exhalation hose 42 in addition to the inspiratory valve 20 already mentioned, and the metabolized breathing gas passes from the lungs of the patient 12 to the atmosphere through the exhalation hose. It is discharged inside.

  The intake hose 32 is connected to the intake valve 20, and the exhalation hose 42 is connected to the exhalation valve 22. Of the two valves, only one of each is open at a time to allow the gas flow to pass. The operation control of the valves 20 and 22 is also performed by the control device 18.

  During the ventilation cycle, during the first inspiratory phase, the exhalation valve 22 is closed and the inhalation valve 20 is opened so that fresh breathing gas can be directed from the housing 14 to the patient 12. The flow of fresh breathing gas is generated by intentionally increasing the pressure of the breathing gas by the pressure changing device 16. Due to the increased pressure, fresh breathing gas flows into the lungs of the patient 12 where the body parts close to the lungs, in particular the thorax, are inflated against the individual elasticity of the body parts close to the lungs. Thereby, the gas pressure inside the lungs of the patient 12 also increases.

  At the end of the inspiratory phase, the inspiratory valve 20 is closed and the expiratory valve 22 is opened. The exhalation phase begins. Due to the gas pressure of the breathing gas in the lungs of the patient 12 rising by the end of the inspiration phase, the breathing gas flows into the atmosphere after the opening of the exhalation valve 22 and continues to flow, the gas in the lungs of the patient 12 The pressure drops. When the gas pressure in the lung 12 reaches the positive end expiratory pressure adjusted in the ventilator 10, i.e., slightly higher than atmospheric pressure, the expiration phase is terminated by the closing of the exhalation valve 22, and further ventilation cycles Will continue.

  During the inspiration phase, the patient 12 is supplied with a so-called tidal volume, i.e. a breathing gas volume per breath. Multiplying the tidal volume of artificial respiration by the number of artificial respiration cycles per minute, that is, the frequency of artificial respiration, gives the minute ventilation of artificial respiration performed here.

  Preferably, the ventilator 10, in particular the controller 18, is configured to repeatedly update or detect the ventilator operating parameters characterizing the ventilator action of the ventilator 10 during the ventilator action. This ensures that the ventilation movements are optimally matched to the patient 12 to be ventilated at each point in time as optimally as possible. Particularly advantageously, the determination of the one or more ventilation operating parameters is carried out using the ventilation frequency, so that for each ventilation cycle, the ventilation operating parameters that are up-to-date and thus optimally adapted to the patient 12 are determined. Can be supplied.

  For this purpose, the ventilator 10 is connected to one or more sensors so as to be able to communicate data, which monitor the condition of the patient and / or the operation of the ventilator. .

  One of these sensors is a proximal flow sensor 44, which detects the dominant respiratory gas flow within the ventilator conduit assembly 30 within the Y piece 45. The flow sensor 44 may be coupled to the data input 26 of the controller 18 using a sensor conduit assembly 46. The sensor conduit assembly 46 may include, but does not necessarily include, an electrical signal transmission line. The sensor conduit assembly 46 may also have hose wires, which transmit the dominant gas pressure on both sides of the flow sensor 44 in the flow direction to the data input 26 for data input. In the unit 26, the gas pressure is quantified by a pressure sensor not shown in FIG. The flow sensor 44 is here represented as a pressure differential flow sensor 44. The flow sensor 44 is preferably a flow sensor that functions on the principle of pressure difference, but may be a flow sensor that functions on the basis of another physical operation principle.

  A further flow sensor 48 is provided in the housing 14 and is labeled as a distal flow sensor 48 because the flow sensor 48 is further away from the patient 12 as compared to the proximal flow sensor 44. .

  The distal flow sensor 48 and its measurement signal may be used to determine the respiratory gas flow through the ventilator conduit assembly 30, for example during the inspiration phase, and more precisely for the distal flow sensor 48, the respiratory gas flow through the inspiratory hose 32. Although it can be used for control, the proximal flow sensor 44 and its measurement signal can be used to control the minute ventilation delivered to the patient 12. Accordingly, the distal flow sensor 48 preferably has a faster response behavior than the proximal flow sensor 44. Because the measurement signal of the distal flow sensor 48 is also used to change the breathing gas flow while breathing, the measurement signal of the proximal flow sensor 44 is preferably only considered for each breath. Based on this, there is no change in the breathing gas flow through the Y-piece 45 while breathing.

  Due to the mounting position within the Y piece 45, the proximal flow sensor 44, unlike the distal flow sensor 48, can basically detect the flow of expiratory gas through the expiratory hose 42.

  The correct functioning of the flow sensors 44 and 48 is important for the correct operation of the ventilator 10 and thus for the health of the patient 12.

  What has become apparent during operation is that the proximal flow sensor 48 is very close to the patient 12, so that the proximal flow sensor 48 has a greater risk of error than the distal flow sensor 48. Proximal flow sensor 44, through which expiratory gas also passes, is stronger than distal flow sensor 48 and is subjected to a load of moisture, for example contained in respiratory gas. This is because the distal flow sensor 48 is located upstream of the regulator 38 in the inspiratory direction, as in the example of FIG. 1, so that only substantially dry inspiratory gas passes through the distal flow sensor 48. This is even more true.

  The control device 18 of the ventilator 10 according to the present invention is configured to monitor the operation of a flow sensor assembly formed from a proximal flow sensor 44 and a distal flow sensor 48, thereby providing a flow sensor. It is possible to detect assembly errors early.

  Based on previous experience, the proximal flow sensor 44 has two types of errors that occur at different rates. That is, a spring error that appears quickly within a few seconds and a drift error that appears gradually within a few minutes.

  First, the detection of spring error should be described.

  In FIG. 2, the detection signal of the proximal flow sensor 44 is indicated by reference numeral 52 as “VTProx”, and the detection signal of the distal flow sensor 48 is indicated by reference numeral 54 as “VTServ”. In that case, the temporally discrete signals in the form directly detected by the sensors 44 and 48 are important.

  Basically, these signals can be used to implement error detection according to the present invention. However, it is recommended that these signals be smoothed, particularly in order to increase the certainty of error detection at the proximal flow sensor 44.

  For the detection of spontaneously occurring spring errors, a smoothing method that considers only a few slight measurement signals and does not suppress short-term sensor signal changes is advantageous. To that end, smoothing, for example by forming a weighted moving average over the last 16 measurements, has proven advantageous. In doing so, the measurement signals of the proximal flow sensor 44 and the distal flow sensor 48 are advantageously smoothed with the same measurement method and the same filter coefficients. However, this is not necessarily the case. For different sensor measurement signals, different smoothing methods or the same smoothing method with different filter coefficients can be used. In the case of this weighted moving average, the filter coefficient is a weighting coefficient for each measurement value. This leads to an unfiltered measurement, which can be expressed as follows for the nth breath of the distal flow sensor 48:

a _i = {1, 2, 3, 4, 5, 6, 7, 8, 8, 7, 6, 5, 4, 3, 2, 1}

  In the example described, there is a linear increase from the most recent measurement to the middle measurement, and a linear decrease from the middle measurement to the oldest measurement. The measurement located at is weighted most strongly. Correspondingly, the proximal sensor signal, smoothed for detection of spring error, may be expressed as:

  Here, the weighting coefficients are the same.

In FIG. 2, the smoothed weighted average VT _SurfFilt is indicated by reference numeral 56, and the smoothed weighted average VT _ProxFilt is indicated by reference numeral 58. The x coordinate of the graph of FIG. 2 indicates time in minutes, and the y coordinate indicates volume in milliliters.

The change value important for the error detection of the ventilator according to the invention is particularly advantageously used as the difference between the latest breath value and its most recent breath value, preferably a smoothed value. Are used to form the difference. Accordingly, the change value _{VT ProxDiff} of change value _{VT SerfDiff} and proximal flow sensor 44 of the distal flow sensor 48 may be formed as follows.

  These change values are shown over time in FIG. In so doing, the change value of the distal flow sensor 48 is indicated by reference numeral 60 and the change value of the proximal flow sensor 44 is indicated by reference numeral 62. Again, the x coordinate of the graph of FIG. 3 indicates time in minutes, and the y coordinate indicates volume, in this case the amount of change in milliliters.

  In FIG. 3, a threshold value is represented by reference numeral 64, and when the change value 62 of the proximal flow sensor 44 exceeds the threshold value, the controller 18 initiates the error detection process described at the beginning of the specification. Since the start of the error detection process is associated with the threshold 64 being exceeded, the computing power of the controller 18 is not unnecessarily consumed in the error detection process.

  If the change value 62 of the proximal flow sensor 44 exceeds the threshold value 64 of the change value 62 of the proximal flow sensor 44 during the kth breath, both of the most recent (k-1) th breaths The change values of the flow sensors 44 and 48 are each stored as a reference value for subsequent comparison.

  Thereafter, the sum of the change values is started.

  A sum is formed for each flow sensor 44 and 48, which is the sum of the most recent (n-1) th breaths for the nth breath, including the most recent change in the nth breath. is there.

  In the illustrated embodiment, the controller 18 detects a spring error because the proximal flow sensor 44 reference value is the sum of the proximal flow sensor 44 values.

The ratio to the threshold value exceeds a predetermined reference threshold, for example, exceeds a reference threshold of 0.08, since the sum of the proximal flow sensor changes since the start of the sum, at least the reference value It means 8%. The reference threshold may have a value different from 8%, but experiments so far have shown that 8% is a very good reference threshold.

  A further condition that arises in addition to the conditions listed above is that the total value of the distal flow sensor 48 is more than an additional reference threshold, which can be, for example, 10% and differs from the total value of the proximal flow sensor 44. That is not.

  Instead of, or in addition to, the additional conditions listed in the previous paragraph, the distal, preferably smoothed, value of the distal flow sensor 48 is greater than another reference threshold value, and the reference A condition can arise that does not differ from the value. This further reference threshold may be, for example, between 1% and 5%. Experiments have shown that yet another baseline threshold of 2% is productive.

  FIG. 6 schematically illustrates the case of a spring error occurring in the proximal flow sensor 44.

  Only when all of the predetermined conditions for reliable detection of the spring error are fulfilled, the control device 18 detects the presence of the spring error and, for example, issues a corresponding alarm.

  Similar smoothed measurements are used to detect gradually occurring drift errors, and preferably smoothing of the measurements is performed on a larger number of past measurements than in the case of spring errors. Is called. For example, the smoothing can also be done, for example, through the formation of a weighted moving average over the last 100 measurements.

At that time, the weighting coefficient b _i is selected so that the sum of all the weighting coefficients is 1. Alternatively or additionally, these measurements can be smoothed through a higher order digital filter, preferably at least a fourth order digital filter.

  In FIG. 4, the raw signal of the distal flow sensor 48 is represented by reference numeral 66, and the raw signal of the proximal flow sensor 44 is represented by reference numeral 68. In contrast, the smoothed signal of the distal flow sensor 48 is represented by reference numeral 70 and the smoothed signal of the proximal flow sensor 44 is represented by reference numeral 72.

  The formation of the change values of the measurement signals 66 and 68 of the distal or proximal flow sensor 48 or 44 is as described above, ie, each sensor's smoothed measurement for the most recent breath and each of the most recent breaths. This is done through the formation of a difference between the sensor's smoothed measurements. That is, the following is effective again.

  In FIG. 5, the change value of the measurement signal of the distal flow sensor 48 is represented by reference numeral 74, and the change value of the measurement signal of the proximal flow sensor 44 is represented by reference numeral 76.

  Again, error detection by the controller 18 may be performed after a predetermined start condition is met, for example, one of the change values leaves the tolerance range near zero, indicated by reference numeral 78 in FIG. Or it starts for the first time when the change values have different signs. Although the upper limit of the tolerance range is not shown in FIG. 5, the tolerance range ranges from −3 ml to +3 ml. Leaving the change value tolerance range may be a starting condition for error detection by the control device 18, but the tolerance range is preferably arranged symmetrically around the zero point. However, this is not necessarily the case, and it is possible to have boundaries with different values depending on the application.

  If there is a predetermined starting condition for detection of drift errors, the controller 18 again begins to form a total value for the change value of each flow sensor 44 or 48 in the manner described above.

  Similarly, the respiration measurements taken by both sensors 44 and 48 immediately before respiration when at least one of the error detection start conditions is met are stored as reference values for subsequent comparisons.

  Here, the error detection of the drift error is not different from the error detection of the spring error.

  In this example, the controller 18 detects the presence of a drift error in the flow sensor assembly because it is the change total value of the proximal flow sensor 44.

The reference value

When the ratio to is above a predetermined reference threshold, and simultaneously, the total change of the distal flow sensor 48

The reference value

This is the case when the ratio to does not exceed a predetermined further reference threshold. The proximal threshold of the proximal flow sensor 44 for detecting drift errors is again 8%, and the further reference threshold may be 2%, for example.

  In this example, controller 18 similarly detects the presence of a drift error in the flow sensor assembly because it is the change total value of distal flow sensor 48.

The reference value

If the ratio to is above a predetermined reference threshold, and at the same time, the total change of the proximal flow sensor 44

The reference value

This is the case when the ratio to does not exceed a predetermined further reference threshold. In doing so, the exemplary values for the reference threshold described in the preceding paragraph are valid.

  Since drift errors can affect either the proximal flow sensor or the distal flow sensor, the conditions for detection of drift errors are preferably symmetric with respect to both reference thresholds.

  As an additional condition for detection of drift error, it can occur that one of both change sum values exceeds a predetermined threshold change amount, for example 2 ml.

  The conditions for detection of drift error using the total change value are schematically shown in FIGS. 7a and 7b. If a drift error is detected, the controller 18 preferably issues an alarm, but the flow sensor assembly preferably continues to operate further.

  In the ventilator 10 according to the present invention, errors in the flow sensor assembly can be detected in a wider range and more accurately than in the past.

DESCRIPTION OF SYMBOLS 10 Ventilator 12 Patient 13 Unit 14 Housing 16 Pressure change apparatus, artificial respiration gas source 18 Control apparatus 20 Intake valve 22 Exhalation valve 24 Data input part 26 Sensor connection part 28 Output apparatus 30 Respiratory conduit assembly 32 Inhalation hose 34 1st Inhalation partial hose 36 Second inhalation partial hose 38 Regulator 40 Liquid container 42 Expiratory hose 44 Proximal flow sensor 45 Y piece 46 Sensor conduit assembly 48 Distal flow sensor 52 Proximity flow sensor 44 detection signal VTPProx
54 Detection signal VTServ of distal flow sensor 48
56 Smoothed Weighted Average VTServFilt
58 Smoothed weighted average VTProxFilt
60 Change value of distal flow sensor 48 62 Change value of proximal flow sensor 44 64 Threshold 66 Raw signal of distal flow sensor 48 68 Raw signal of proximal flow sensor 44 70 Smoothing of distal flow sensor 48 Signal 72 Smoothed signal of proximal flow sensor 44 74 Measurement signal change value of distal flow sensor 48 76 Measurement signal change value of proximal flow sensor 44 78 Tolerance range

Claims (15)

A ventilator (10) for performing artificial respiration,
-Artificial ventilation gas source (16),
-A ventilation conduit assembly (30) extending between said ventilation gas source (16) and the proximal end on the patient side;
A valve assembly (20, 22) comprising an inhalation valve (20) and an exhalation valve (22);
A flow sensor assembly (44, 48) for quantitative detection of gas flow in the artificial respiration conduit assembly (30), which is located far from the patient end of the artificial respiration conduit assembly (30); A flow sensor assembly having a distal flow sensor (48) configured and a proximal flow sensor (44) located near a patient end of the ventilator conduit assembly (30); and
A control device (18) for processing at least the measurement signals of the flow sensor assembly (44, 48), the measurement signals of the distal flow sensor (48) and / or the proximal flow sensor (44); A controller (18) configured to infer an error based on
In a ventilator (10) having:
The control device (18)
The change values (62, 76) of the measurement signals (54, 58, 68, 72) of the proximal flow sensor (44),
-Comparison with the change values (60, 74) of the measurement signals (52, 56, 66, 70) of the distal flow sensor (48) and / or
-Based on comparison with measurement signals (54, 58, 68, 72) of said proximal flow sensor (44) and / or
The change values (60, 74) of the measurement signals (52, 56, 66, 70) of the distal flow sensor (48),
-Comparison with the change values (62, 76) of the measurement signals (54, 58, 68, 72) of the proximal flow sensor (44) and / or
-Based on a comparison with the measurement signal (52, 56, 66, 70) of said distal flow sensor (48),
A ventilator configured to infer errors in the flow sensor assembly (44, 48).   Of the measurement signal (54, 58, 68, 72 or 52, 56, 66, 70) of one of the distal flow sensor (48) and the proximal flow sensor (44). The change value (62, 76 or 60, 74) has the opposite change direction from the change value (60, 74 or 62, 76, 60, 74) of the other flow sensor (48 or 44), respectively. The ventilator according to claim 1, characterized in that the control device (18) is configured to infer an error of the flow sensor assembly (44, 48) in some cases. A first value of the sum of the change values (62, 76), which occur in succession, preferably in direct succession, of the measurement signals (54, 58, 68, 72) of the proximal flow sensor (44); The ratio to the reference value exceeds a predetermined reference threshold; and
A second value of the sum of the change values (60, 74), which occur in succession, preferably directly in succession, of the measurement signal (52, 56, 66, 70) of the distal flow sensor (48); When the ratio to the reference value is below a predetermined reference threshold,
Respirator according to claim 1 or 2, characterized in that the control device (18) is configured to infer errors in the flow sensor assembly (44, 48).   The first reference value and / or the second reference value are measured signal values of the proximal flow sensor (44), in particular the start of the sum of the change values (62, 76, 60, 74). Respirator according to claim 3, characterized in that it is the measured signal value of the proximal flow sensor (44) at the time or just before the start.   To start summing the change values (62, 76, 60, 74) when the measured signal value and / or change value of the proximal flow sensor (44) exceeds a predetermined signal threshold value (64), Respirator according to claim 3 or 4, characterized in that the control device (18) is configured.   The controller (18) is configured to process the measurement signals of the proximal flow sensor (44) and the distal flow sensor (48) for each breath, preferably the proximal flow sensor ( The proximal flow, wherein a first and / or second reference value is assigned to the breath immediately before the threshold breath when the measured signal value of 44) exceeds a predetermined signal threshold (64) in the threshold breath Respirator according to claim 5, characterized in that it is a measured signal value of the sensor (48). If the ratio of the measured value of the proximal flow sensor (44), which occurs in succession, preferably the sum of consecutive changes, to the third reference value exceeds a predetermined reference threshold, as well as,
The measurement signal of the distal flow sensor (48), if the ratio of the sum of the change values that occur in succession, preferably in direct succession, to the fourth reference value is below a predetermined reference threshold value. Inferring an error in the flow sensor assembly (44, 48), or
If the ratio of the measurement value of the proximal flow sensor (44), which occurs in succession, preferably the sum of the successive changes, to the fifth reference value is below a predetermined reference threshold, as well as,
The ratio of the sum of the change values of the measurement signals of the distal flow sensor (48), which occur in succession, preferably in direct succession, to a sixth reference value exceeds a predetermined reference threshold value. Inferring errors in the flow sensor assembly (44, 48),
Respirator according to any one of the preceding claims, characterized in that the control device (18) is configured. In addition,
The sum of the change values, preferably in direct succession, of the measurement signal of the proximal flow sensor (44), which is in succession, exceeds a predetermined first sum threshold, and / or
The flow sensor assembly only if the sum of the change values, which occur in succession, preferably in direct succession, of the measurement signal of the distal flow sensor (48) exceeds a predetermined second total threshold value; The respirator according to claim 7, characterized in that the control device (18) is configured to infer an error of (44, 48).   The third reference value and / or the fifth reference value is a measurement signal value of the proximal flow sensor (44), in particular the proximal at the beginning or just before the start of the sum of the change values. 9. Ventilator according to claim 7 or 8, characterized in that it is a measurement signal value of the flow sensor (44).   The fourth reference value and / or the sixth reference value is a measurement signal value of the distal flow sensor (48), in particular the distal at the start or just before the start of the sum of the change values. 10. A ventilator according to any one of claims 7 to 9, characterized in that it is a measurement signal value of the flow sensor (48).   A change value for forming a total value when the value of the change value of the measurement signal of the proximal flow sensor (44) and / or the distal flow sensor (48) exceeds a predetermined threshold value (78). The ventilator according to any one of claims 7 to 10, characterized in that the sum of   The controller (18) is adapted to filter the measurement signal of the flow sensor assembly (44, 48) by filtering, in particular by forming a moving average, preferably by forming a moving weighted average, or by a higher order digital It is configured to smooth using a filter, preferably a digital filter of higher order than 2, particularly preferably a fourth order digital filter, and the control device (18) is preferably further smoothed. The ventilator according to any one of claims 1 to 11, wherein the ventilator is configured to detect a change value based on the measured signal.   The proximal flow sensor (44) and the distal flow sensor (48) each provide one, preferably exactly one measurement value representing the gas flow assigned to the breath for each breath. The ventilator according to any one of claims 1 to 12.   14. A ventilator according to any one of the preceding claims, characterized in that the proximal flow sensor (44) is a pressure difference sensor.   The controller (18) is configured to issue an alarm when an error is inferred, preferably the processing of the measurement signal of the proximal flow sensor (44) is continued. The ventilator according to any one of claims 1 to 14.